Fix (rag): epub load

bot.go

@@ -830,11 +830,21 @@ func chatWatcher(ctx context.Context) {
 func showSpinner() {
 	spinners := []string{"⠋", "⠙", "⠹", "⠸", "⠼", "⠴", "⠦", "⠧", "⠇", "⠏"}
 	var i int
-	for botRespMode {
+	botPersona := cfg.AssistantRole
+	if cfg.WriteNextMsgAsCompletionAgent != "" {
+		botPersona = cfg.WriteNextMsgAsCompletionAgent
+	}
+	for botRespMode || toolRunningMode {
 		time.Sleep(100 * time.Millisecond)
 		spin := i % len(spinners)
 		app.QueueUpdateDraw(func() {
-			textArea.SetTitle(spinners[spin] + " input")
+			if toolRunningMode {
+				textArea.SetTitle(spinners[spin] + " tool")
+			} else if botRespMode {
+				textArea.SetTitle(spinners[spin] + " " + botPersona)
+			} else {
+				textArea.SetTitle(spinners[spin] + " input")
+			}
 		})
 		i++
 	}
@@ -1219,7 +1229,9 @@ func findCall(msg, toolCall string) bool {
 	}
 	// Show tool call progress indicator before execution
 	fmt.Fprintf(textView, "\n[yellow::i][tool: %s...][-:-:-]", fc.Name)
+	toolRunningMode = true
 	resp := callToolWithAgent(fc.Name, fc.Args)
+	toolRunningMode = false
 	toolMsg := string(resp) // Remove the "tool response: " prefix and %+v formatting
 	logger.Info("llm used a tool call", "tool_name", fc.Name, "too_args", fc.Args, "id", fc.ID, "tool_resp", toolMsg)
 	fmt.Fprintf(textView, "%s[-:-:b](%d) <%s>: [-:-:-]\n%s\n",

go.mod

@@ -6,17 +6,19 @@ require (
 	github.com/BurntSushi/toml v1.5.0
 	github.com/GrailFinder/google-translate-tts v0.1.3
 	github.com/GrailFinder/searchagent v0.2.0
+	github.com/PuerkitoBio/goquery v1.11.0
 	github.com/gdamore/tcell/v2 v2.13.2
 	github.com/glebarez/go-sqlite v1.22.0
 	github.com/gopxl/beep/v2 v2.1.1
 	github.com/gordonklaus/portaudio v0.0.0-20250206071425-98a94950218b
 	github.com/jmoiron/sqlx v1.4.0
+	github.com/ledongthuc/pdf v0.0.0-20250511090121-5959a4027728
 	github.com/neurosnap/sentences v1.1.2
 	github.com/rivo/tview v0.42.0
+	github.com/yuin/goldmark v1.4.13
 )
 
 require (
-	github.com/PuerkitoBio/goquery v1.11.0 // indirect
 	github.com/andybalholm/cascadia v1.3.3 // indirect
 	github.com/dustin/go-humanize v1.0.1 // indirect
 	github.com/ebitengine/oto/v3 v3.4.0 // indirect

go.sum

@@ -43,6 +43,8 @@ github.com/hashicorp/golang-lru/v2 v2.0.7 h1:a+bsQ5rvGLjzHuww6tVxozPZFVghXaHOwFs
 github.com/hashicorp/golang-lru/v2 v2.0.7/go.mod h1:QeFd9opnmA6QUJc5vARoKUSoFhyfM2/ZepoAG6RGpeM=
 github.com/jmoiron/sqlx v1.4.0 h1:1PLqN7S1UYp5t4SrVVnt4nUVNemrDAtxlulVe+Qgm3o=
 github.com/jmoiron/sqlx v1.4.0/go.mod h1:ZrZ7UsYB/weZdl2Bxg6jCRO9c3YHl8r3ahlKmRT4JLY=
+github.com/ledongthuc/pdf v0.0.0-20250511090121-5959a4027728 h1:QwWKgMY28TAXaDl+ExRDqGQltzXqN/xypdKP86niVn8=
+github.com/ledongthuc/pdf v0.0.0-20250511090121-5959a4027728/go.mod h1:1fEHWurg7pvf5SG6XNE5Q8UZmOwex51Mkx3SLhrW5B4=
 github.com/lib/pq v1.10.9 h1:YXG7RB+JIjhP29X+OtkiDnYaXQwpS4JEWq7dtCCRUEw=
 github.com/lib/pq v1.10.9/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=
 github.com/lucasb-eyer/go-colorful v1.3.0 h1:2/yBRLdWBZKrf7gB40FoiKfAWYQ0lqNcbuQwVHXptag=
@@ -67,6 +69,7 @@ github.com/rivo/uniseg v0.4.7 h1:WUdvkW8uEhrYfLC4ZzdpI2ztxP1I582+49Oc5Mq64VQ=
 github.com/rivo/uniseg v0.4.7/go.mod h1:FN3SvrM+Zdj16jyLfmOkMNblXMcoc8DfTHruCPUcx88=
 github.com/stretchr/testify v1.10.0 h1:Xv5erBjTwe/5IxqUQTdXv5kgmIvbHo3QQyRwhJsOfJA=
 github.com/stretchr/testify v1.10.0/go.mod h1:r2ic/lqez/lEtzL7wO/rwa5dbSLXVDPFyf8C91i36aY=
+github.com/yuin/goldmark v1.4.13 h1:fVcFKWvrslecOb/tg+Cc05dkeYx540o0FuFt3nUVDoE=
 github.com/yuin/goldmark v1.4.13/go.mod h1:6yULJ656Px+3vBD8DxQVa3kxgyrAnzto9xy5taEt/CY=
 golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
 golang.org/x/crypto v0.0.0-20210921155107-089bfa567519/go.mod h1:GvvjBRRGRdwPK5ydBHafDWAxML/pGHZbMvKqRZ5+Abc=

main.go

@@ -7,6 +7,7 @@ import (
 var (
 	boolColors        = map[bool]string{true: "green", false: "red"}
 	botRespMode       = false
+	toolRunningMode   = false
 	editMode          = false
 	roleEditMode      = false
 	injectRole        = true

rag/extractors.go

@@ -0,0 +1,184 @@
+package rag
+
+import (
+	"archive/zip"
+	"bytes"
+	"errors"
+	"fmt"
+	"io"
+	"os"
+	"os/exec"
+	"path"
+	"strings"
+
+	"github.com/PuerkitoBio/goquery"
+	"github.com/ledongthuc/pdf"
+	"github.com/yuin/goldmark"
+	"github.com/yuin/goldmark/extension"
+	"github.com/yuin/goldmark/parser"
+	"github.com/yuin/goldmark/renderer/html"
+)
+
+func ExtractText(fpath string) (string, error) {
+	ext := strings.ToLower(path.Ext(fpath))
+	switch ext {
+	case ".txt":
+		return extractTextFromFile(fpath)
+	case ".md", ".markdown":
+		return extractTextFromMarkdown(fpath)
+	case ".html", ".htm":
+		return extractTextFromHtmlFile(fpath)
+	case ".epub":
+		return extractTextFromEpub(fpath)
+	case ".pdf":
+		return extractTextFromPdf(fpath)
+	default:
+		return "", fmt.Errorf("unsupported file format: %s", ext)
+	}
+}
+
+func extractTextFromFile(fpath string) (string, error) {
+	data, err := os.ReadFile(fpath)
+	if err != nil {
+		return "", err
+	}
+	return string(data), nil
+}
+
+func extractTextFromHtmlFile(fpath string) (string, error) {
+	data, err := os.ReadFile(fpath)
+	if err != nil {
+		return "", err
+	}
+	return extractTextFromHtmlContent(data)
+}
+
+// non utf-8 encoding?
+func extractTextFromHtmlContent(data []byte) (string, error) {
+	doc, err := goquery.NewDocumentFromReader(bytes.NewReader(data))
+	if err != nil {
+		return "", err
+	}
+	// Remove script and style tags
+	doc.Find("script, style, noscript").Each(func(i int, s *goquery.Selection) {
+		s.Remove()
+	})
+	// Get text and clean it
+	text := doc.Text()
+	// Collapse all whitespace (newlines, tabs, multiple spaces) into single spaces
+	cleaned := strings.Join(strings.Fields(text), " ")
+	return cleaned, nil
+}
+
+func extractTextFromMarkdown(fpath string) (string, error) {
+	data, err := os.ReadFile(fpath)
+	if err != nil {
+		return "", err
+	}
+	// Convert markdown to HTML
+	md := goldmark.New(
+		goldmark.WithExtensions(extension.GFM),
+		goldmark.WithParserOptions(parser.WithAutoHeadingID()),
+		goldmark.WithRendererOptions(html.WithUnsafe()), // allow raw HTML if needed
+	)
+	var buf bytes.Buffer
+	if err := md.Convert(data, &buf); err != nil {
+		return "", err
+	}
+	// Now extract text from the resulting HTML (using goquery or similar)
+	return extractTextFromHtmlContent(buf.Bytes())
+}
+
+func extractTextFromEpub(fpath string) (string, error) {
+	r, err := zip.OpenReader(fpath)
+	if err != nil {
+		return "", fmt.Errorf("failed to open epub: %w", err)
+	}
+	defer r.Close()
+
+	var sb strings.Builder
+
+	for _, f := range r.File {
+		ext := strings.ToLower(path.Ext(f.Name))
+		if ext != ".xhtml" && ext != ".html" && ext != ".htm" && ext != ".xml" {
+			continue
+		}
+
+		// Skip manifest, toc, ncx files - they don't contain book content
+		nameLower := strings.ToLower(f.Name)
+		if strings.Contains(nameLower, "toc") || strings.Contains(nameLower, "nav") ||
+			strings.Contains(nameLower, "manifest") || strings.Contains(nameLower, ".opf") ||
+			strings.HasSuffix(nameLower, ".ncx") {
+			continue
+		}
+
+		rc, err := f.Open()
+		if err != nil {
+			continue
+		}
+
+		if sb.Len() > 0 {
+			sb.WriteString("\n\n")
+		}
+		sb.WriteString(f.Name)
+		sb.WriteString("\n")
+
+		buf, readErr := io.ReadAll(rc)
+		rc.Close()
+		if readErr == nil {
+			sb.WriteString(stripHTML(string(buf)))
+		}
+	}
+
+	if sb.Len() == 0 {
+		return "", errors.New("no content extracted from epub")
+	}
+	return sb.String(), nil
+}
+
+func stripHTML(html string) string {
+	var sb strings.Builder
+	inTag := false
+	for _, r := range html {
+		switch r {
+		case '<':
+			inTag = true
+		case '>':
+			inTag = false
+		default:
+			if !inTag {
+				sb.WriteRune(r)
+			}
+		}
+	}
+	return sb.String()
+}
+
+func extractTextFromPdf(fpath string) (string, error) {
+	_, err := exec.LookPath("pdftotext")
+	if err == nil {
+		out, err := exec.Command("pdftotext", "-layout", fpath, "-").Output()
+		if err == nil && len(out) > 0 {
+			return string(out), nil
+		}
+	}
+	return extractTextFromPdfPureGo(fpath)
+}
+
+func extractTextFromPdfPureGo(fpath string) (string, error) {
+	df, r, err := pdf.Open(fpath)
+	if err != nil {
+		return "", fmt.Errorf("failed to open pdf: %w", err)
+	}
+	defer df.Close()
+	textReader, err := r.GetPlainText()
+	if err != nil {
+		return "", fmt.Errorf("failed to extract text from pdf: %w", err)
+	}
+	var buf bytes.Buffer
+	_, err = io.Copy(&buf, textReader)
+	if err != nil {
+		return "", fmt.Errorf("failed to read pdf text: %w", err)
+	}
+	return buf.String(), nil
+}

rag/rag.go

@@ -7,7 +7,6 @@ import (
 	"gf-lt/models"
 	"gf-lt/storage"
 	"log/slog"
-	"os"
 	"path"
 	"regexp"
 	"sort"
@@ -58,7 +57,7 @@ func wordCounter(sentence string) int {
 func (r *RAG) LoadRAG(fpath string) error {
 	r.mu.Lock()
 	defer r.mu.Unlock()
-	data, err := os.ReadFile(fpath)
+	fileText, err := ExtractText(fpath)
 	if err != nil {
 		return err
 	}
@@ -68,7 +67,6 @@ func (r *RAG) LoadRAG(fpath string) error {
 	default:
 		r.logger.Warn("LongJobStatusCh channel is full or closed, dropping status message", "message", LoadedFileRAGStatus)
 	}
-	fileText := string(data)
 	tokenizer, err := english.NewSentenceTokenizer(nil)
 	if err != nil {
 		return err

tables.go

@@ -247,9 +247,49 @@ func formatSize(size int64) string {
 	return fmt.Sprintf("%.1f%s", s, units[i])
 }
 
-func makeRAGTable(fileList []string) *tview.Flex {
-	actions := []string{"load", "delete"}
-	rows, cols := len(fileList), len(actions)+2
+type ragFileInfo struct {
+	name     string
+	inRAGDir bool
+	isLoaded bool
+	fullPath string
+}
+
+func makeRAGTable(fileList []string, loadedFiles []string) *tview.Flex {
+	// Build set of loaded files for quick lookup
+	loadedSet := make(map[string]bool)
+	for _, f := range loadedFiles {
+		loadedSet[f] = true
+	}
+
+	// Build merged list: files from ragdir + orphaned files from DB
+	ragFiles := make([]ragFileInfo, 0, len(fileList)+len(loadedFiles))
+	seen := make(map[string]bool)
+
+	// Add files from ragdir
+	for _, f := range fileList {
+		ragFiles = append(ragFiles, ragFileInfo{
+			name:     f,
+			inRAGDir: true,
+			isLoaded: loadedSet[f],
+			fullPath: path.Join(cfg.RAGDir, f),
+		})
+		seen[f] = true
+	}
+
+	// Add orphaned files (in DB but not in ragdir)
+	for _, f := range loadedFiles {
+		if !seen[f] {
+			ragFiles = append(ragFiles, ragFileInfo{
+				name:     f,
+				inRAGDir: false,
+				isLoaded: true,
+				fullPath: "",
+			})
+		}
+	}
+
+	rows := len(ragFiles)
+	cols := 4 // File Name | Preview | Action | Delete
 	fileTable := tview.NewTable().
 		SetBorders(true)
 	longStatusView := tview.NewTextView()
@@ -273,7 +313,7 @@ func makeRAGTable(fileList []string) *tview.Flex {
 			SetAlign(tview.AlignCenter).
 			SetSelectable(false))
 	fileTable.SetCell(0, 2,
-		tview.NewTableCell("Load").
+		tview.NewTableCell("Load/Unload").
 			SetTextColor(tcell.ColorWhite).
 			SetAlign(tview.AlignCenter).
 			SetSelectable(false))
@@ -284,18 +324,29 @@ func makeRAGTable(fileList []string) *tview.Flex {
 			SetSelectable(false))
 	// Add the file rows starting from row 1
 	for r := 0; r < rows; r++ {
+		f := ragFiles[r]
 		for c := 0; c < cols; c++ {
 			color := tcell.ColorWhite
 			switch {
 			case c == 0:
+				displayName := f.name
+				if !f.inRAGDir {
+					displayName = f.name + " (orphaned)"
+				}
 				fileTable.SetCell(r+1, c,
-					tview.NewTableCell(fileList[r]).
+					tview.NewTableCell(displayName).
 						SetTextColor(color).
 						SetAlign(tview.AlignCenter).
 						SetSelectable(false))
 			case c == 1:
-				fpath := path.Join(cfg.RAGDir, fileList[r])
-				if fi, err := os.Stat(fpath); err == nil {
+				if !f.inRAGDir {
+					// Orphaned file - no preview available
+					fileTable.SetCell(r+1, c,
+						tview.NewTableCell("not in ragdir").
+							SetTextColor(tcell.ColorYellow).
+							SetAlign(tview.AlignCenter).
+							SetSelectable(false))
+				} else if fi, err := os.Stat(f.fullPath); err == nil {
 					size := fi.Size()
 					modTime := fi.ModTime()
 					preview := fmt.Sprintf("%s | %s", formatSize(size), modTime.Format("2006-01-02 15:04"))
@@ -312,15 +363,32 @@ func makeRAGTable(fileList []string) *tview.Flex {
 							SetSelectable(false))
 				}
 			case c == 2:
+				actionText := "load"
+				if f.isLoaded {
+					actionText = "unload"
+				}
+				if !f.inRAGDir {
+					// Orphaned file - can only unload
+					actionText = "unload"
+				}
 				fileTable.SetCell(r+1, c,
-					tview.NewTableCell("load").
-						SetTextColor(color).
-						SetAlign(tview.AlignCenter))
-			default:
-				fileTable.SetCell(r+1, c,
-					tview.NewTableCell("delete").
+					tview.NewTableCell(actionText).
 						SetTextColor(color).
 						SetAlign(tview.AlignCenter))
+			case c == 3:
+				if !f.inRAGDir {
+					// Orphaned file - cannot delete from ragdir (not there)
+					fileTable.SetCell(r+1, c,
+						tview.NewTableCell("-").
+							SetTextColor(tcell.ColorDarkGray).
+							SetAlign(tview.AlignCenter).
+							SetSelectable(false))
+				} else {
+					fileTable.SetCell(r+1, c,
+						tview.NewTableCell("delete").
+							SetTextColor(color).
+							SetAlign(tview.AlignCenter))
+				}
 			}
 		}
 	}
@@ -376,12 +444,16 @@ func makeRAGTable(fileList []string) *tview.Flex {
 			pages.RemovePage(RAGPage)
 			return
 		}
-		// For file rows, get the filename (row index - 1 because of the exit row at index 0)
-		fpath := fileList[row-1] // -1 to account for the exit row at index 0
-		// notification := fmt.Sprintf("chat: %s; action: %s", fpath, tc.Text)
+		// For file rows, get the file info (row index - 1 because of the exit row at index 0)
+		f := ragFiles[row-1]
+		// Handle "-" case (orphaned file with no delete option)
+		if tc.Text == "-" {
+			pages.RemovePage(RAGPage)
+			return
+		}
 		switch tc.Text {
 		case "load":
-			fpath = path.Join(cfg.RAGDir, fpath)
+			fpath := path.Join(cfg.RAGDir, f.name)
 			longStatusView.SetText("clicked load")
 			go func() {
 				if err := ragger.LoadRAG(fpath); err != nil {
@@ -398,8 +470,25 @@ func makeRAGTable(fileList []string) *tview.Flex {
 				})
 			}()
 			return
+		case "unload":
+			longStatusView.SetText("clicked unload")
+			go func() {
+				if err := ragger.RemoveFile(f.name); err != nil {
+					logger.Error("failed to unload file from RAG", "filename", f.name, "error", err)
+					_ = notifyUser("RAG", "failed to unload file; error: "+err.Error())
+					app.QueueUpdate(func() {
+						pages.RemovePage(RAGPage)
+					})
+					return
+				}
+				_ = notifyUser("RAG", "file unloaded successfully")
+				app.QueueUpdate(func() {
+					pages.RemovePage(RAGPage)
+				})
+			}()
+			return
 		case "delete":
-			fpath = path.Join(cfg.RAGDir, fpath)
+			fpath := path.Join(cfg.RAGDir, f.name)
 			if err := os.Remove(fpath); err != nil {
 				logger.Error("failed to delete file", "filename", fpath, "error", err)
 				return

tui.go

@@ -913,6 +913,7 @@ func init() {
 					return nil
 				}
 			}
+			// Get files from ragdir
 			fileList := []string{}
 			for _, f := range files {
 				if f.IsDir() {
@@ -920,22 +921,14 @@ func init() {
 				}
 				fileList = append(fileList, f.Name())
 			}
-			chatRAGTable := makeRAGTable(fileList)
-			pages.AddPage(RAGPage, chatRAGTable, true, true)
-			return nil
-		}
-		if event.Key() == tcell.KeyCtrlY { // Use Ctrl+Y to list loaded RAG files
-			// List files already loaded into the RAG system
-			fileList, err := ragger.ListLoaded()
+			// Get loaded files from vector DB
+			loadedFiles, err := ragger.ListLoaded()
 			if err != nil {
 				logger.Error("failed to list loaded RAG files", "error", err)
-				if notifyerr := notifyUser("failed to list RAG files", err.Error()); notifyerr != nil {
-					logger.Error("failed to send notification", "error", notifyerr)
-				}
-				return nil
+				loadedFiles = []string{} // Continue with empty list on error
 			}
-			chatLoadedRAGTable := makeLoadedRAGTable(fileList)
-			pages.AddPage(RAGLoadedPage, chatLoadedRAGTable, true, true)
+			chatRAGTable := makeRAGTable(fileList, loadedFiles)
+			pages.AddPage(RAGPage, chatRAGTable, true, true)
 			return nil
 		}
 		if event.Key() == tcell.KeyRune && event.Modifiers() == tcell.ModAlt && event.Rune() == '1' {